The development of strategies to manipulate T-cell immunity for the treatment of human cancer, has proceeded rapidly during the last decade and yielded encouraging results, but obstacles hindering progress and broader application have been identified. Adoptive transfer of large number of tumor-reactive cytolytic T cells has often yielded incomplete responses, in part due to the antigenic target specificities and to functional limitations of the infused cells. However, failure to detect, in most patients, T cells reactive to tumors which can be manipulated for therapeutic benefit remains the major hurdle. Most defined tumor antigens are not uniquely expressed in the tumor, but rather can also be detected in normal tissues, and the generation of T cells reactive with such tumor- associated antigens must overcome the obstacles of tolerance to self-proteins and potential consequences of auto-reactivity if the tolerance can be broken. The proposed studies will utilize cellular and molecular techniques, including retroviral gene transfer as a means to modify T-cell function, and transgenic (TG) mouse models developed by the investigator to examine immunity to tumor- associated antigens. The ultimate goals of this project are to improve the efficacy of therapy in settings in which T cells can be identified, and to enhance the likelihood of obtaining tumor- reactive T cells in settings in which this has been difficult. The proposed specific aims are: 1) To determine if T cells can be rendered tumor-reactive and effective in tumor therapy by the introduction via retroviral vectors of T-cell receptor (TCR) genes; 2) To determine if CD8+ CTL can be genetically modified to provide an autocrine growth signal and function independently of CD4+ Th by introducing chimeric cytokine receptors; 3) To determine in TG mice expressing a tumor antigen in normal tissues the autoimmune consequences of CD8+ T cell tumor therapy; 4) To determine if enhanced presentation of antigen by conditioned dendritic cells and enhanced costimulatory signals can overcome tolerance to a tumor- associated antigen; and 5) To determine the properties that distinguish tolerant T cells from responding T cells employing both biochemical techniques and analysis of gene expression with microarray, and asses the ability of strategies designed to overcome tolerance to correct these abnormalities.